CN111353950B

CN111353950B - Method for image processing, medical imaging device and electronically readable data carrier

Info

Publication number: CN111353950B
Application number: CN201911299221.2A
Authority: CN
Inventors: B.施赖伯
Original assignee: Siemens Healthineers AG
Current assignee: Siemens Medical Ag
Priority date: 2018-12-20
Filing date: 2019-12-17
Publication date: 2024-02-27
Anticipated expiration: 2039-12-17
Also published as: US11599977B2; US20200202501A1; CN111353950A; DE102018222595A1

Abstract

Method for image processing of an image dataset (1) of a patient imaged with a medical imaging device (9), in particular an X-ray device, wherein the image dataset (1) has image values associated with image points and shows an imaging region of the patient with at least one enhanced object, in particular a medical apparatus (7), which object is mapped by image values within an image value interval, wherein the method comprises the following steps: determining a non-linear high pass filtered enhancement data set, the enhancement data set being limited to an image portion comprising image values lying within an image value interval; determining a result data set (8) by adding enhancement data sets weighted with weighting values to the image data set (1); and outputting the result data set (8).

Description

Method for image processing, medical imaging device and electronically readable data carrier

Technical Field

The invention relates to a method for image processing of an image data record of a patient recorded with a medical imaging device, in particular an X-ray device, wherein the image data record has image values associated with image points and shows a recording region of the patient with at least one enhanced object, in particular a medical device, which object is mapped by image values within image value intervals. The invention further relates to a medical imaging device, in particular an X-ray device, a computer program and an electronically readable data carrier.

Background

Medical imaging, such as X-ray imaging, and increasingly other modalities, such as magnetic resonance imaging, are used in order to be able to assess the progress and/or success of a surgical procedure, in particular a minimally invasive procedure, on a patient. In such pre-, inter-, intra-and/or post-operative imaging, for example, it is necessary to evaluate: the extent to which the medical device, in particular the implant and/or the instrument, is correctly positioned in order to achieve the desired medical effect and/or to determine the desired diagnostic information.

An example of such a surgical procedure, in particular a minimally invasive procedure, is the placement of a stent in a blood vessel of a patient. During or after placement of the stent, the respective physician must ensure that the placement of the stent is correct. This can become difficult, in particular, if the stent is close to other medical devices, such as metal coils, and/or surrounded by strongly weakened anatomical structures, such as bones. It can therefore generally be said that it is always difficult to embody a medical image data set of a recording region of a patient in which a specific object, in particular a medical device, is to be evaluated when approaching other anatomical structures and/or medical aids.

Disclosure of Invention

The object of the present invention is therefore to create a possibility for image processing which makes it possible to achieve an improved visualization of high-contrast objects, in particular medical devices, in a medical image data set.

The technical problem is solved by a method, a medical imaging device, a computer program and an electronically readable data carrier according to the invention. The invention also provides an advantageous design scheme.

The method of the initially mentioned type according to the invention comprises the following steps:

-determining a non-linear high-pass filtered enhancement data set, the enhancement data set being limited to an image portion comprising image values lying within an image value interval;

-determining a result data set by adding the enhancement data set weighted with the weighting values to the image data set; and

-outputting the result data set.

It is thereby proposed to use nonlinear filtering, which produces an enhancement data set, which, when added to the original medical image data set, results in a result data set in which the object to be evaluated, in particular the medical device, i.e. for example the implant and/or the instrument, can be identified extremely clearly without enhancing the noise level and/or likewise surrounding objects and/or generating artifacts. For this purpose, on the one hand, it is provided that the filtering is applied only to the following regions of the medical image data set: in this region, there are image values in a predefined image value range which is associated with the object to be enhanced and thus describes in which region, for example in the case of X-ray imaging, there are image values in the gray-scale region and/or in the HU value region which display the image points, i.e. pixels or voxels, of the object to be enhanced. Here, the image value interval may be defined for a high contrast object such as a medical device as follows: the image value interval excludes anatomical structures in the region of the shot and/or does not comprise typical image values of other medical auxiliary devices that may be located in the region of the shot. The nonlinear filtering is thereby applied specifically to image regions which form or as precisely as possible comprise the image portion which displays the object to be enhanced.

The definition of the image value intervals can be based on theoretical considerations, in particular in connection with the material properties of the object important for imaging, and/or on previous measurements of the object, wherein in particular also targeted calibration measurements can be performed.

The enhancement data set here preferably contains only image portions which have been non-linearly filtered, wherein it is also conceivable in an embodiment to implement at least initially also further regions which are subsequently removed.

Here, the specific possibility rule for determining the enhancement data set is carried out by the following steps:

-determining the intermediate data set by applying a nonlinear low-pass filter to image points satisfying a selection condition for evaluating the image values, wherein the selection condition selects an image point when the image value of the image point lies within an interval of image values, and

-determining the enhancement data set by subtracting the intermediate data set image point by image point from the image data set.

In the framework of the invention, on the one hand, it is shown that low-pass filters which require the required characteristics (which will be described in more detail later) have been specifically proposed in the prior art and can therefore also be applied in the framework of the invention. However, in the case of the described determination of the enhancement data set by subtraction, it is particularly advantageous if the image regions of the medical image data set which are not filtered and therefore do not show the object to be enhanced are removed directly, so that a direct enhancement can be achieved by subsequently adding the enhancement data set to the image data set. It is therefore provided that a nonlinear low-pass filter is applied to the medical image data set, which may generally be two-dimensional or preferably three-dimensional. In this case, it is also possible for the nonlinear low-pass filter to be applied only to image points whose image values lie in a predefined image value range, the image value range containing image values which are typical for high-contrast objects to be enhanced. In the case of a medical device as object to be reinforced, this means that, in particular, no filtering of anatomical structures such as soft tissue, bones or air is performed, which is also applicable to other medical aids or the like which generate image values lying outside the image value interval. But these unfiltered image points remain first in the intermediate data set, since after filtering a non-linear high-pass filtered enhancement data set is determined by correspondingly subtracting the low-pass filtered intermediate data set from the original medical image data set.

Here, it is essential in the framework of the present invention that nonlinear filtering is performed. In this case, a particularly preferred embodiment of the invention provides that for the low-pass filtering, the image points located around the observed image points within the filter mask used are evaluated as a function of the difference between the image values of the image points located around the observed image points and the image values of the observed image points, wherein in particular image values with a large deviation lead to a weaker weighting, and the filtering process is carried out as a function of the evaluation. Furthermore, the use of bilateral filters and/or weighted median filters as low-pass filters is particularly advantageous, since the two mentioned low-pass filter types already have a corresponding weighting set as a function of the image value differences.

That is to say, in particular in the framework of non-linearities of the filter application, intensity weighting is carried out which penalizes excessive image value differences with respect to neighboring values, for example image value differences which significantly exceed the standard deviation which is substantially given by noise, resulting in that the corresponding image values are not taken into account in the low-pass filtering or in significantly smaller weights. In this way it is achieved that voxels containing the object to be enhanced do not deteriorate, for example by other objects, for example other medical aids in the vicinity acquired by the filter mask used. It has been shown that particularly pronounced enhancement of objects to be enhanced can be achieved in this way, while at the same time amplification of noise effects can be largely avoided and that other objects, in particular adjacent other objects, such as anatomical structures and/or medical aids, are neither amplified nor cause a deterioration of the image impression, for example due to filtered artifacts.

A further embodiment of the invention regarding the filtering process may also provide that, in particular as an additional selection condition, image points of the individual deviations are detected, which lie in the region of the image values in the image value interval, whereas the image values of these image points lie outside the image value interval and the image points of the individual deviations are likewise filtered. It can thus be monitored whether non-characteristic points (ausreiβer) occur in the region or regions of the image where the image values are usually located within the interval of the image values (such that the region is associated with the object to be enhanced), although non-characteristic points can also be included in the filtering, which in particular also contributes to the smoothing of the image look and feel.

In this case, according to the invention, the object may preferably be a medical device, in particular a stent. Although stents are in principle high contrast objects, stents are still generally mapped weaker than other medical aids (e.g. coils introduced into aneurysms), especially in cases where the wall thickness is quite small. The method according to the invention is therefore particularly advantageous, in particular for evaluating stents, since it has been shown that stents can be more clearly identified in the result data set. While not increasing the noise level of the soft tissue region nor enhancing the bone structure, as with other medical aids, such as coils inserted into aneurysms, generally result in significantly higher intensities and thus significantly higher image values. It can therefore generally be said that the image value intervals are advantageously selected such that anatomical structures in the recording region that do not correspond to the subject and/or coils in medical aids that do not correspond to the subject, in particular aneurysms, are not selected for filtering. Different types of catheters have also been proposed as further examples of stents for use as medical devices.

The weighting values are of course coefficients that are to be selected to be greater than zero, the weighting values determining the intensity of the enhancement, and the weighting values being suitably selected such that a significant enhancement of the object to be enhanced occurs in the resulting data set without excessively changing the image look and feel, so that the image may also be represented in its entirety. The weighting values can be selected, for example, in the range of 1 to 20, in particular in the range of 5 to 10, and/or can be set, for example, by means of a controller on the user side.

Preferably, the image data set may be three-dimensional and may exist as a cross-sectional image or a layer image. The determination of the associated result image of the result data set can then be performed continuously for all sectional images or layer images. In particular, it is then conceivable to limit the image processing to a part of the entire data set when required, for example when the enhancement should be performed only in a specific sectional image or layer image.

Although it is conceivable and advantageous to store the result data set for further use and/or to transmit it to a further computing device for the purpose of outputting the result data set, it may in particular be provided that the result data set is displayed in particular in a volume-rendered manner and/or as a particularly thin maximum intensity projection and/or as a multi-planar reconstruction. The enhancement effect produced by performing the steps of the method according to the invention shows a significant improvement of the visibility of the object to be enhanced for views of Volume Rendering (VRT), maximum Intensity Projection (MIPs), thin maximum intensity projection (MMIP) and multi-planar reconstruction (MRP).

In addition to the method, the invention also relates to a medical imaging device having a control device designed to perform the method according to the invention. All embodiments of the method according to the invention can be similarly transferred to the medical imaging device according to the invention, so that the already mentioned advantages can be achieved with the medical imaging device according to the invention. In particular, the medical imaging device may be an X-ray device, for example an X-ray device with a C-arm, on which the X-ray source and the X-ray detector are oppositely arranged. Such a C-arm X-ray device may be used particularly advantageously in surgical procedures, in particular in minimally invasive procedures, on patients in order to monitor the progress of the procedure and/or to determine the success of the procedure. In particular, for an X-ray apparatus with a C-arm, it is also conceivable to record projection images from different projection directions, for example during rotation of the C-arm, in order to thereby obtain a basis from which a three-dimensional image data set of the recorded region can be reconstructed. However, the use of other X-ray devices, such as computed tomography devices, for applications in the case of medical interventions has been proposed; recording intraoperative or postoperative medical image data sets has also been proposed in connection with magnetic resonance apparatuses.

The control device may in particular have at least one processor and at least one memory component. In particular, the control device may implement a plurality of functional units to perform the steps of the present invention. The control device using the medical imaging apparatus, in particular for use in surgery, has the following advantages: the enhancement can be performed directly and the result data set can be displayed, for example, on at least one display device of the medical imaging apparatus, in particular a monitor, wherein the display device is advantageously located in a position visible from the surgical site of the doctor. However, it is also conceivable to execute the method according to the invention on other computing devices, for example on a computing device of a workstation or a viewing station.

In particular, a control device or a computing device designed for carrying out the method according to the invention may have a filter unit for determining the enhancement data set, a determination unit for determining the result data set, and an output unit, in particular an output interface, for outputting the result data set. The filter unit here comprises in particular a low-pass filter subunit and a subtraction subunit as subunits. The filter unit and the determination unit may be image processors.

The medical imaging device may also have an input device, by means of which, for example, a class of the object to be enhanced may be selected and then associated with the respective selection parameter and/or filter parameter. The selection parameters and/or filter parameters comprise, for example, parameters describing the image value interval, parameters describing the filter mask, parameters describing the allowable deviation (i.e. allowable image value differences) with respect to adjacent image points and/or weighting values. The latter can also be selected separately if necessary, in particular during the display of the result data set, so that the user can select a display form suitable for himself.

The computer program according to the invention can be loaded directly into a memory means of a computing means, in particular of a control means of a medical imaging device, for example, and has program means for performing the steps of the method according to the invention when the computer program is executed in the computing means. The computer program may be stored on an electronically readable data carrier according to the invention, which thus comprises electronically readable control information stored thereon, which electronically readable control information comprises at least one of the mentioned computer programs and is designed to perform the method according to the invention when the data carrier is used in a computing device. The data carrier may in particular be a non-transitory data carrier, such as a CD-ROM.

Drawings

Further advantages and details of the invention are derived from the embodiments described below and with reference to the drawings. In the accompanying drawings:

figure 1 shows a flow chart of an embodiment of the method according to the invention,

figure 2 shows a schematic view of a medical image data set,

figure 3 shows a schematic diagram of the resulting data set,

fig. 4 shows a medical imaging device according to the invention, and

fig. 5 shows a functional structure of a control device of the medical imaging apparatus.

Detailed Description

Fig. 1 shows a flow chart of an embodiment of the method according to the invention. In this case, the progress or success is evaluated when the stent is placed in the blood vessel of the patient, for which purpose in step S1 a corresponding medical image data record is acquired intraoperatively or postoperatively using a medical imaging device, in this case an X-ray device with a C-arm. In this case, projection images are acquired from different projection directions and a three-dimensional medical image data set is reconstructed, which may be present as a sectional image or as a slice image.

Fig. 2 shows a schematic illustration of such a medical image data record 1. The rough course 3 of anatomical structures, for example bones 2 and blood vessels, is represented with low intensity or low image values, which are difficult to identify, i.e. at the respective image points. The metallic coil 5 inserted into the aneurysm 4 can be seen clearly. On the contrary, the stent 6, which is visible but whose contrast may not be completely or sufficiently accurately acquired, appears very blurred as the medical device 7 to be reinforced.

Thus, in step S2, the user selects via the input means of the medical imaging device that the display stand 6 should be enhanced in the medical image data set 1. The stent 6 or other object to be enhanced (which may furthermore be an anatomical structure such as a bone) is here associated with certain selection parameters and filter parameters, which are now more precisely determined and parameterized for the subsequent image processing process, so that an enhanced display of the stent 6 is formed in the resulting data set.

For this purpose, in step S3, a nonlinear low-pass filter, in particular a bilateral filter or a weighted median filter, is applied to the image points of the image data set 1 belonging to an image value interval which describes the usual display of the support 6 and is therefore associated with the support 6 as selection parameter and filter parameter. For this purpose, for example, a selection criterion can be used, which selects, as the image points to be filtered, the image points whose image values lie in the image value interval. Furthermore, the further selection criteria can also detect and likewise select individual non-characteristic points (ausreiβer), that is to say image points lying within the region of the image values in the image value interval whose image values lie outside the image value interval individually. The selected image points are then filtered through a nonlinear low pass filter.

In both examples of the low-pass filter mentioned, the nonlinearity of the low-pass filter is here embodied in that the image values of the other pixels, which lie adjacent to one another in the filter mask of the low-pass filter, are heavily deviated from the image values of the image points to be filtered, are weighted less heavily in the filtering. In principle, such image values and their image points can be ignored or still be processed with a smaller weighting in the filtering. As a measure for the allowable deviation, a standard deviation defined by noise may be used herein. In this way, in particular the influence of adjacent, severely deviating structures, for example metallic structures such as coils 5, on the filtering can be avoided, so that artifacts or false enhancements can also be reduced in this respect. Furthermore, noise amplification is thereby at least reduced.

The result of the low-pass filtering at the selected image points is an intermediate data set which now also comprises the unfiltered part of the image data set 1. In step S4, the intermediate data set is used to determine the enhancement data set by subtracting the intermediate data set from the image data set 1. Since in particular the entire region of the support 6 not containing the object to be reinforced is not filtered, said entire region is completely removed, so that only the reinforced support 6 remains in the reinforced data set in its as precise a position and size range as possible.

In step S5, a result data set is generated by adding an enhancement data set, which is weighted by a weighting value, i.e. a coefficient greater than zero, to the image data set 1. The result data set thus determined may be output in step S6.

Fig. 3 shows a schematic diagram of such a result data set 8. It can be clearly seen there that the clearly defined and contrast-improving stent 6 can be seen more clearly as a medical device 7 to be reinforced with respect to the anatomy and in particular with respect to the coils in the aneurysm 4.

In particular, the output of the result data set can be displayed by means of a display device of the medical imaging device, for example by means of a monitor, in addition to the storage in particular. Here, outputs such as VRT, MIP, thin MIP, MPR, and the like are conceivable.

Fig. 4 shows an embodiment of a medical imaging device 9 according to the invention, which is designed here as an X-ray device with a C-arm 10 on which an X-ray source 11 and an X-ray detector 12 are arranged opposite each other. The medical imaging device 9 is suitable for surgical operation and is thus associated with an operating table 13.

The operation of the X-ray device 9 is controlled by a control device 14 which is also designed to perform the method according to the invention. For displaying the result data set a display device 15, for example a monitor, may be used. User input may be made through the operating device 16.

Fig. 5 shows the functional structure of the control device 14 in more detail. In addition to the camera unit 17 controlling the camera operation of the imaging device 9 in a manner known in principle, the control device has a filter unit 18 for carrying out the steps S3 and S4, the filter unit 18 in turn having a low-pass filter subunit 19 for carrying out the step S3 and a subtraction subunit 20 for carrying out the step S4. The enhancement data set determined in the filter unit 18 is used in the determination unit 21 for performing step S5 and thus for determining a result data set, which can be displayed, for example, on the display device 15 via the output unit 22 (step S6).

In particular, the filter unit 18 and the determination unit 21 are image processors, while the output unit 22 may be or may comprise an output interface.

Although the invention has been illustrated and described in detail with reference to preferred embodiments, the invention is not limited to the examples disclosed and other variations can be derived therefrom by a person skilled in the art without departing from the scope of protection of the invention.

Claims

1. Method for image processing of an image dataset (1) of a patient imaged with a medical imaging device (9), wherein the image dataset (1) has image values associated with image points and shows an imaged region of the patient with at least one enhanced object, which object is mapped by image values within a predefined image value interval, wherein the method comprises the following steps:

determining a non-linear high-pass filtered enhancement data set, said enhancement data set being limited to image portions comprising image values lying within an interval of image values,

-generating a result data set (8) by directly enhancing the image data set with a non-linear high-pass filtered enhancement data set, wherein the non-linear high-pass filtered enhancement data set is added to the image data set (1), and wherein the non-linear high-pass filtered enhancement data set is weighted by a weighting value, the weighting value comprising a value larger than zero, the weighting value defining the intensity of the non-linear high-pass filtered enhancement data set; and

outputting the result data set (8),

wherein the following steps are performed for determining the enhanced data set:

-determining the intermediate data set by applying a nonlinear low-pass filter to image points meeting a selection condition for evaluating image values, wherein the selection condition selects image points when their image values lie within an image value interval, and

-determining the enhancement data set by subtracting the intermediate data set image point by image point from the image data set (1).

2. The method according to claim 1, characterized in that the object is a medical device (7).

3. Method according to claim 1, characterized in that for low-pass filtering, image points located around the observed image points within the filter mask used are evaluated on the basis of the difference between the image values of the image points located around the observed image points and the image values of the observed image points, and that the filtering process is performed on the basis of the evaluation; and/or use bilateral filters and/or weighted median filters as low pass filters.

4. The method according to claim 2, characterized in that the medical device (7) is a stent (6); and/or selecting the image value interval such that anatomical structures in the imaging region that do not correspond to the subject and/or medical auxiliary devices that do not correspond to the subject are not selected for filtering.

5. A method according to any one of claims 1 to 3, characterized in that the weighting value is selected in the range of 1 to 20.

6. A method according to any one of claims 1 to 3, characterized in that the image data set (1) is three-dimensional and exists as a sectional image or layer image, wherein the determination of the associated result image of the result data set (8) is performed continuously for all sectional images or layer images.

7. A method according to any of claims 1 to 3, characterized in that the result data set (8) is displayed in a volume-rendered manner and/or as a maximum intensity projection and/or as a multi-planar reconstruction.

8. Medical imaging device (9) having a control means (14) designed for performing the method according to any of the preceding claims.

9. An electronically readable data carrier, having stored thereon a computer program which, when executed in a computing device, performs the steps of the method according to any of claims 1 to 7.